Structural Optimization Design of Radial Tire Building Machine based on Radial Based Function Neural Network and NSGA- II

In order to ensure the lightweight design of radial tire building machine under the condition of not less than the initial dynamic and static performance, finite element analysis, topology optimization, approximate model combined with NSGA-II and other methods are used to carry out multi-objective optimization design of the whole machine, taking mass and first-order natural frequency as objective functions. Firstly, the finite element analysis and modal experiment are combined to ensure the accuracy of the finite element model and the pedestal is determined to be the key structural component affecting the dynamic performance of the whole machine. Then, topology optimization is carried out for the pedestal, and the optimal basic structure is redesigned according to the material distribution. Through sensitivity analysis, the key structural dimensions sensitive to the objective function are selected as design variables. Latin hypercube experimental design is used to conduct sample design and randomly generate 81 groups of design schemes. Based on the simulation data of design schemes, the RBF neural network approximation model of universities is established. Finally, NSGA-II is used for multi-objective optimization, and the optimal design variable group is obtained. Under the condition that the dynamic and static performance of radial tire building machine is not reduced, the mass is reduced by 17.9%, and the first-order natural frequency is increased by 19.7%.